Skip to main content
GET
Search runs

Authorizations

Authorization
string
header
required

Bearer authentication header of the form Bearer <token>, where <token> is your auth token.

Headers

X-Organization-Id
string | null

Query Parameters

evaluation_issue_category
string | null

Filter to runs whose current evaluation (completed and not superseded) has at least one evaluation_issue with this category. Superseded or non-completed evaluations are ignored, so a retried run whose latest evaluation passed will not match.

evaluation_issue_subcategory
string | null

Filter to runs whose current evaluation (completed and not superseded) has at least one evaluation_issue with this subcategory. When combined with evaluation_issue_category both must match the same issue row. Superseded or non-completed evaluations are ignored.

evaluation_status
enum<string> | null

Filter by the run's current evaluation verdict. passed / failed require a completed non-superseded evaluation with overall_passed set to true / false respectively (three-valued logic — neither matches a NULL verdict). not_evaluated matches the UI's "Not analyzed" badge: either no completed non-superseded evaluation exists (no-eval, pending, judge-errored, or superseded-only) OR the current evaluation completed without emitting a verdict (overall_passed IS NULL). Combined with evaluation_issue_category / evaluation_issue_subcategory, both predicates resolve against the same evaluation row (by the data invariant that at most one current evaluation exists per run).

Available options:
passed,
failed,
not_evaluated
q
string | null

ANTLR filter expression

limit
integer | null

Page size

Required range: 1 <= x <= 500
sort_by
string | null

Sort expression (e.g. created_at:desc)

fields
string | null

Field projection (e.g. id,name)

embed
string | null

Relation embedding. Available values: comfyui_workflow, current_evaluation_summary, flow, logs, model, node_runs. Use a comma-separated list to request multiple embeds. Protected: current_evaluation_summary requires evaluations:read.

pq
string | null

Pagination cursor query

from_prev
boolean
default:false

Backward pagination flag

Response

Paginated run list.

Paginated list response.

items
FullValidator · object[]
required
first
string | null
next
string | null
previous
string | null